Is something rotten in the state of "Update"?

[Gremlin]

Another terminal message when executing reconfig of Active-Backup Bond
{6.0.650 - active-backup bond did not survive reboot}

Code: Select all

bond0: option fail_over_mac: unable to set because the bond device has slaves

[TMzethar]

It seems you are wondering why it prioritizes 1000Mbps over 2500Mbps in Active-backup Bond mode.
This is automatically selected by the Bond tool, and as you guessed, it may have chosen the first enabled LAN.
In our general products, we have never had a 'high speed+low speed' network card configuration, and therefore, there has been no related optimization design.

[Gremlin]

Well, you guys built it that way Also guess no one expected 221 buyers to have 2.5Gbe networks.
Isn't progress wonderful.
I had hoped that Active-backup would provide some redundancy via the slower link but I have given up for now. Just too much hassle.
Gone back to the old way of doing things with fixed IPs, a mapped 'network disk' (or disks) (using TNAS-PC for setup) and 2 live nic for when necessary.

Will hope multichannel smb provides some benefit. (I have increased nic 'RSS' value on desktop as well.) Everything is aggravated by some connections having to traverse an aggregated (trunk) link which is not necessarily a good idea (apparently).

Just for info:
a) My -424 also uses LAN2 as the link channel but as both LAN are 2.5Gbe it does not appear so obvious and has no detrimental effect that I can see.

b) Working over the aggregated (trunk) link it sometimes takes an inordinate amount of time for windows (11 in particular case) to recognise a device name on Samba. It can be seen in windows explorer but cannot authenticate.